www.nature.com/scientificreports
OPEN
received: 19 March 2015 accepted: 17 November 2015 Published: 16 December 2015
TALENs-mediated gene disruption of FLT3 in leukemia cells: Using genome-editing approach for exploring the molecular basis of gene abnormality Jue Wang1,*, Tongjuan Li1,*, Mi Zhou1, Zheng Hu2, Xiaoxi Zhou1, Shiqiu Zhou1, Na Wang1, Liang Huang1, Lei Zhao1, Yang Cao1, Min Xiao1, Ding Ma2, Pengfei Zhou3, Zhen Shang1 & Jianfeng Zhou1,2 Novel analytic tools are needed to elucidate the molecular basis of leukemia-relevant gene mutations in the post-genome era. We generated isogenic leukemia cell clones in which the FLT3 gene was disrupted in a single allele using TALENs. Isogenic clones with mono-allelic disrupted FLT3 were compared to an isogenic wild-type control clone and parental leukemia cells for transcriptional expression, downstream FLT3 signaling and proliferation capacity. The global gene expression profiles of mutant K562 clones and corresponding wild-type controls were compared using RNA-seq. The transcriptional levels and the ligand-dependent autophosphorylation of FLT3 were decreased in the mutant clones. TALENs-mediated FLT3 haplo-insufficiency impaired cell proliferation and colony formation in vitro. These inhibitory effects were maintained in vivo, improving the survival of NOD/SCID mice transplanted with mutant K562 clones. Cluster analysis revealed that the gene expression pattern of isogenic clones was determined by the FLT3 mutant status rather than the deviation among individual isogenic clones. Differentially expressed genes between the mutant and wild-type clones revealed an activation of nonsense-mediated decay pathway in mutant K562 clones as well as an inhibited FLT3 signaling. Our data support that this genome-editing approach is a robust and generally applicable platform to explore the molecular bases of gene mutations. Acute leukemia (AL) is among the most common forms of hematological malignancies and exhibits a striking heterogeneity in genetic makeup and clinical outcome. Despite striking success in treating some subtypes of AL due to progress in the understanding of cytogenetic/gene mutations and the introduction of corresponding targeted therapies, many genetically high-risk AL subtypes remain refractory to current standard regimens and display poor prognosis, as the current therapeutic strategies are insufficient to cure genetically high-risk AL1,2. The increasing classification/stratification of AL by the World Health Organization (WHO) and National Comprehensive Cancer Network (NCCN) based on genetic orientation highlights the profound impact of cytogenetic/gene mutations on decision-making regarding clinical management. Therefore, there is an urgent need to dissect the molecular basis of AL in these patients and to develop novel treatment strategies. The advances in high-throughput sequencing technologies enabled the characterization of AL from a global genomic perspective in an unbiased and comprehensive manner rather than a candidate gene approach. The identification of highly recurrent gene mutations in AL has provided invaluable preclinical rationales for improved diagnosis, patient stratification and targeted therapy3,4. Nevertheless, although exhaustive landscapes of AL genomes are emerging, elucidating how these recurrent gene mutations are associated with clinical phenotypes and treatment 1
Department of hematology, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Hubei, China. 2Cancer Biology Research Center, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Hubei, China. 3Wuhan YZY Bio-Pharma Co., Ltd., Wuhan, Hubei, China. *These authors contributed equally to this work. Correspondence and requests for materials should be addressed to Z.S. (email: [email protected]) or J.Z. (email: [email protected]) Scientific Reports | 5:18454 | DOI: 10.1038/srep18454
1
www.nature.com/scientificreports/ outcomes has proven difficult. The ability to determine the clinical relevance of these identified individual genetic abnormalities is greatly needed. More importantly, uncovering the molecular events that are central to leukemia biology is a major challenge in the human genome era and is fundamentally important for the future of genetic medicine5. Several approaches have been widely used to explore the molecular mechanism underlying a given gene mutation. Many elegant leukemia animal models have been established using genetically engineered strategies and have provided persuasive in vivo insight into the driving factors of leukemogenesis. In the coming years, genetically engineered animal models will remain among the most reliable tools to study the association between genetic defects and clinical phenotypes, and contribution to the design and development of novel molecular-targeted strategies. Nevertheless, the animal model approach has some intrinsic shortcomings. For example, animal models may not always accurately mimic the leukemia phenotype relevant to the clinical setting6, aside from the time-consuming and expensive process of generating and maintaining a transgenic model. Alternatively, cell lines and primary sample-based models have also been widely used to explore the critical molecular events that lead to tumor cell phenotypes. Several factors have limited the identification of actual molecular mechanisms due to the huge discrepancy in genetic backgrounds among primary samples, the non-physiological levels of transgene overexpression/knockdown and the interference introduced by randomly integrated viral vectors. Therefore, novel analytic tools are urgently needed to address the molecular mechanisms underlying leukemia-relevant gene function in the post-genome era. Transcription activator-like effector (TALE) nucleases (TALENs), an efficient genome editing tool, are artificial fusion proteins containing the catalytic domain of the endonuclease FokI and a designed TALE DNA-binding domain that recognizes a specific DNA sequence7,8. The binding of two separate TALENs to adjacent DNA sequences enables dimerization of FokI and cleavage of the target DNA, introducing site-specific double-strand breaks (DSBs). Subsequently, cellular DNA repair via either homology-directed repair or the non-homologous end joining (NHEJ) pathway is activated9. Thus far, genome-editing technology has been successfully applied to induce myeloid malignancy in normal hematopoietic stem cells in mice10. However, few studies have performed this technique to investigate the molecular events caused by individual gene abnormality in leukemia. In this study, we generated isogenic clones, in two individual leukemia cell lines, by disrupting FMS-like tyrosine kinase 3 (FLT3) gene in a single allele using designed TALENs. The resulting isogenic clones which were only different in the FLT3 mutant status were compared for FLT3 downstream signaling, proliferation capacity and transcriptional expression. Our data strongly support that this genome-editing approach can serve as a robust and generally applicable platform for exploring the molecular basis of a given gene abnormality.
Results
Generation of isogenic leukemia clones carrying disrupted FLT3 in a single allele using designed TALENs. To generate isogenic leukemia clones carrying a disrupted FLT3 juxtamembrane (JM) domain in a
single allele, a pair of TALENs targeting exon14 of FLT3 was designed. The TALEN target site was selected within the sequence of exon14, which encodes the JM domain and is located upstream of the tyrosine kinase domains (TKDs) and the kinase insert (KI) domain, such that insertions or deletions caused by NHEJ could result in disruption of the reading frame or the formation of a premature stop codon. We constructed TALENs composed of 17.5 and 15.5 repeats to cleave a site using a 15 bp spacer according to five computationally derived design guidelines as described previously (Fig. 1a)11. Since the transfection efficiency in K562 cells approached over 90% (Supplementary Fig. S1) under the experimental conditions described in METHODS, the T7E1 mismatch sensitive assay was applied to assess the nuclease activity of the designed TALENs at their intended target in K562 cells. As shown in Fig.1b, T7E1 cleavage products of the expected size were detected on both the electrophoretic trace and the pseudo-gel image, displaying an estimated frequency of NHEJ events of up to 6%. These data demonstrated robust genomic editing capacity of the designed TALENs at the JM domain of endogenous FLT3 in leukemia cells, facilitating subsequent screening of leukemia clones carrying genetically disrupted FLT3. Before establishing isogenic clonal leukemia models, we assessed the basal expression level of FLT3 in a panel of working cell lines representing AML and CML myeloid blast crisis by Western blot (Supplementary Fig. S2). Consistent with previous studies12,13, FLT3 protein levels in cell lines harboring FLT3-ITD mutations such as MV4-11 and MOLM-13 were generally higher than those cell lines with wild-type FLT3, whereas FLT3 wild-type cell lines expressed FLT3 protein at different levels. Based on transfection efficiencies and cell viabilities after nucleofection, we chose three individual cell lines, K562 and OCI-AML3 with the lowest and the highest basal wild-type FLT3 expression respectively, and Jurkat cells with undetectable FLT3 expression as negative control14, to generate isogenic leukemia clones using our designed TALENs. Immunoprecipitation followed by mass spectrometry (Supplementary methods) was performed to address the specificity of FLT3 protein bands in K562 cells (Supplementary Fig. S3 and Supplementary Fig. S4). PCR and Sanger sequencing were conducted to confirm that there was no FLT3-ITD mutation in the three working parental cell lines. The TALEN-nucleofected leukemia cells were seeded at low density in a 96-well plate for recovery and homogeneous expansion. The clonal cell populations were isolated and screened via PCR amplification of genomic DNA and Sanger sequencing to identify indels characteristic of NHEJ (Fig. 1c). In this experiment, among a total of 384 K562 clones screened, 11 clones (~2.86%) contained heterozygous modifications in exon 14, including 5 clones carrying frameshift mutations and 6 clones carrying in-frame deletions in one allele (Fig. 1d). All 5 isogenic K562 clones carrying a frameshift mutation in FLT3 were predicted to contain a premature termination codon, of which three clones (clones k20, k114 and k324) were selected for the subsequent experiments. Similar results were obtained from OCI-AML3 and Jurkat, and 3 clones from OCI-AML3 (clones o35, o166 and o253) and 2 clones from Jurkat (clones j105 and j212) were confirmed with frame shift mutations (Supplementary Table S1)
Scientific Reports | 5:18454 | DOI: 10.1038/srep18454
2
www.nature.com/scientificreports/
a
JM ECD
KI
TM
TKD1
TALEN L exon14
exon13
Protein
TKD2
exon15
Genome
TALEN R primer F
primer R
483 bp
CTTCTACGTTGATTTCAGAGAATATGAATATGATCTCAAATGGGAGTTTCCAAGAGAAA TALEN target GAAGATGCAACTAAAGTCTCTTATACTTATACTAGAGTTTACCCTCAAAGGTTCTCTTT
b
Fraction cleaved:12.5%
bp 500
Estimated %NHEJ:6.46%
Fluorescence Units[x1E3]
20
c 66,455
K562+control 479bp
66,465
Clone k20 wt allele 66,475 66,485
400
10 0 20 10
300
309bp
Clone k20 mut allele
K562+TALENs 479bp 309bp 163bp
0
d
479bp
200 160
163bp
exon 14
100 Clone#
Exon14 sequence
wt k20 k57 k86 k114 k195 k221 k273 k279 k324 k359 k366
CTTCTACGTTGATTTCAGAGAATATGAATATGATCTCAAATGGGAGTTTCCAAGAGAAA CTTCTACGTTGATTTC-------------------TCAAATGGGAGTTTCCAAGAGAAA CTTCTACGTTGATTTCAGA------GAATATGATCTCAAATGGGAGTTTCCAAGAGAAA CTTCTACGTTGATTTCAGA------GAATATGATCTCAAATGGGAGTTTCCAAGAGAAA CTTCTACGTTGATTTCAGAGAATATGA-----ATCTCAAATGGGAGTTTCCAAGAGAAA CTTCTACGTTGATTTCAGA------GAATATGATCTCAAATGGGAGTTTCCAAGAGAAA CTTCTACGTTGATTTCAGAGAATATGAA---------AAATGGGAGTTTCCAAGAGAAA CTTCTACGTTGATTTCAGAGAATATG------------AATGGGAGTTTCCAAGAGAAA CTTCTACGTTGATTTCAGA------GAATATGATCTCAAATGGGAGTTTCCAAGAGAAA CTTCTACGTTGATTTCAGAGAATATG-----------AAATGGGAGTTTCCAAGAGAAA CTTCTACGTTGATTTCAGAGAATATCAATTATGATCTCAAATGGGAGTTTCCAAGAGAA CTTCTACGTTGATTTCAGAGAAT----------------ATGGGAGTTTCCAAGAGAAA
Frame shift 19 6 6 5 6 9 12 6 11 1 16
NHEJ events ( 11/384=~2.86% )
Figure 1. TALEN-mediated gene disruption of FLT3 in the JM domain in leukemia cells. (a) Schematic overview of the strategy to disrupt the JM domain of FLT3. The three exons that encode the JM domain, the TALEN pair used (TALEN L and TALEN R) and the primer pair (Primer F and Primer R) for PCR and DNA sequencing to screen the mutant clones were shown. The boxes indicate the TALEN binding sequences in the lower diagram in grey. (b) A T7 Endonuclease I assay presented as an electrophoretic trace (the left upper and lower panels) and a pseudo-gel image (the right panel). The blue-colored trace and the lane labeled with a blue bar at the top correspond to K562 cells transfected with the empty TALEN expression vector as a negative control. The red-colored trace and the lane labeled with a red bar at the top correspond to K562 cells transfected with the TALEN pair. The red arrows denote T7EI cleavage products of the appropriate size, and the single black arrow denotes the original uncleaved PCR product in both the electrophoretic trace and the pseudo-gel image. The internal lane standards were labeled with a black bar. (c) Representative sequencing results of targeted K562 clones displaying either the wild-type or targeted mutant allele. The original base numbers in the genomic region of the wild-type FLT3 gene were shown. The dotted lines denote the 19bp deletion in the mutant allele. The colored bars below each chromatogram indicate the predicted proteins. The black bar with an asterisk represents the premature termination codon. (d) Summary of Sanger sequencing results of positive isogenic mutant clones derived from the K562 cell line transfected with TALENs targeting FLT3 exon14. The TALEN binding sequences are shown in bold letters in the wild-type clone. The deletions are indicated by dashes, and insertion or base substitution is denoted by a black triangle or an asterisk, respectively. ECD, extracellular domain; TM, transmembrane domain; JM, juxtamembrane domain; TKD, tyrosine kinase domain; KI, kinase insert; TALEN, transcription activator-like effector nuclease; WT, wild-type; Mut, mutant; Bp, base pair; NHEJ, non-homologous end joining. Scientific Reports | 5:18454 | DOI: 10.1038/srep18454
3
www.nature.com/scientificreports/
Clone analyzed
Putative off-target sites examined
Mutated sites detected (reference sequence)
Comparison to parental K562 sequence
Clone kw1
20
2
Identical (20/20)
Clone k20
20
2
Identical (20/20)
Clone k114
20
2
Identical (20/20)
Clone k324
20
2
Identical (20/20)
Table 1. Summary of the off-target analysis to isogenic clones. The 20 most possible off-target sites identified using the TALE-NT 2.0 Paired Target Finder were sequenced and aligned with either reference sequence (human reference genome, NCBI Build 36.1) or parental K562 sequence.
after screening 480 OCI-AML3 clones and 300 Jurkat clones, respectively. Three wild-type clones derived from the same screening procedure for each cell line were randomly selected as negative controls. To assess the off-target gene effects of the designed TALENs, Paired Target Finder was employed to search for potential off-target sites. The genomic region containing the 20 most likely off-target candidate sites was PCR-amplified and subjected to sequencing analysis. Mutations within two candidate sites were identified in all isogenic K562 subclones examined; however, the parental K562 cells were confirmed to carry the identical mutations (Table 1). Therefore, no off-target events caused by the designed TALENs were detected.
Single-allelic disruption of FLT3 induces FLT3 haplo-insufficiency in Leukemia cells. To deter-
mine whether the premature termination codon caused by NHEJ mutation in the mutant clones of each cell lines induces FLT3 haplo-insufficiency, we first examined the FLT3 mRNA transcript levels in the parental cell line and its isogenic clones via real-time RT-PCR. As expected, significantly lower transcript levels of FLT3 were detected in the mutant clones compared to those of wild-type clones or the parental cell population for both K562 and OCI-AML3 (Fig.2a), implying that premature stop codon-mediated mRNA decay might occur in the mutant FLT3 clones. Next, we sought to determine whether the NHEJ mutation induced a reduced level of FLT3 autophosphorylation upon exposure to treatment of the isogenic mutant clones with FLT3 ligand (FL). Since the K562 cells express the FLT3 protein at a very low level, immunoblotting was performed following immunoprecipitation with FLT3 antibody to examine the potential FL-dependent activity of FLT3 in the K562 clones. As shown in Fig.2b, western blot analyses revealed substantially attenuated FL-dependent autophosphorylation of FLT3 in the mutant clones compared to the wild-type K562 clones, although the reduction of total FLT3 protein level was not significant after immunoprecipitation. The target genes downstream of the FLT3 signaling pathway were further examined via Western blot. Although consistent activation of FLT3 target genes was detected in a FL-dependent manner in all K562 clones, the FL-induced phosphorylation level was substantially reduced in the mutant K562 clones compared to their wild-type counterparts (Fig. 2c). Because K562 cells carry the oncogenic bcr/abl fusion gene, which is also a potent activator of the downstream pathways described above15, we sought to determine whether the TALENs exerted off-target effects on the bcr/abl protein. As shown in Fig. 2d, neither the basal nor phosphorylated level of bcr/abl was affected by the TALENs procedure, indicating that FLT3 haplo-insufficiency was specifically caused by the introduced mutations. A set of similar immunoblotting assays was performed to validate the TALENs-mediated FLT3 haplo-insufficiency for OCI-AML3 clones (Fig. 2e). As results, a significant down-regulation of FLT3 protein expression and ligand-dependent FLT3 autophosphorylation in the mutant OCI-AML3 clones was easily detected using standard immunoblotting, whereas similar trends of weakened Akt, Erk and Stat5 signaling were observed. However, no such alterations regarding the FLT3 protein expression or downstream signaling could be detected in the negative control Jurkat cells and its isogenic clones (Supplementary Fig. S5). To confirm that the changes in cellular signaling were caused by FLT3 gene disruption, siRNA knockdown was performed in K562, OCI-AML3 and THP-1 (Supplementary methods). As shown in Supplementary Fig. S6, similar alterations in FLT3 signaling occurred 48 hours after siRNA transfection. Collectively, these results demonstrate that TALEN-mediated disruption of exon14 of FLT3 induces FLT3 haplo-insufficiency as indicated by reduced FLT3 autophosphorylation and substantially attenuated FLT3 signaling.
TALEN-mediated FLT3 haplo-insufficiency impairs cell proliferation and colony forming capacity in vitro. It is well established that one of the most prominent characteristics of activated FLT3 signaling is the
promotion of cell proliferation16. Therefore, we sought to investigate the biologic effect of TALEN-mediated FLT3 haplo-insufficiency on leukemia cell proliferation in vitro. Based on the CCK8 (Fig. 3a) and CFSE (Fig. 3b) assays in both K562 and OCI-AML3 cells, the proliferation capacity in the mutant FLT3 clones was consistently reduced to nearly 50% or less than those of the corresponding wild-type FLT3 clones, while no inhibition on the proliferation capacity could be observed in the mutant Jurkat clones compared to their wild-type counterparts (Supplementary Fig. S7). In the same experiments, no significant difference in cell proliferation capacity was detected between the isogenic mutant clones (P > 0.05). It was also to our surprise that even in K562 cells that express FLT3 at a very low level, the TALEN-mediated FLT3 haplo-insufficiency resulted in significantly impaired proliferation capacity. To rule out the possibility that TALEN-mediated growth inhibition might be an artefact caused by multiple experimental procedures, we generated and sequence-verified an artificial heterozygous FLT3-ITD model (clone KI-H6) in K562 using the same pair of TALENs by homologous directed repair adopting similar protocols as previously reported (Supplementary Figure S8)17. By comparing growth capacities of three clones with different FLT3 mutant status (FLT3-wt/wt, wt/-, wt/ITD), we found that the FLT3-ITD knock-in clone (KI-H6) display similar proliferative patterns to that of FLT3 wt/wt clones in the presence of serum (Supplementary Figure S8). The findings support that the growth inhibition on FLT3 wt/- clones was specifically due to FLT3 gene disruption Scientific Reports | 5:18454 | DOI: 10.1038/srep18454
4
www.nature.com/scientificreports/ RT-PCR FLT3 1.5
1.5
ns
1.0
1.0
0.5
0.5
p-abl
Bcr-abl β-actin
0.0
Flt3 ligand -
+
-
p-Akt t-Akt p-Erk t-Erk p-Stat5 t-Stat5 GAPDH
+
-
+
-
+
-
-
+
o2 53
o1 66
o3 5
1
cl on e
ow +
cl on e
+
cl on e
Flt3 ligand -
cl on e
O C I-A
e
-
-
+
+
p-Flt3
k3 24
Flt3
cl
on e
k1 14
on e cl
on e cl
on e cl
K5 62
kw
c
k2 0
1
p-Flt3 Flt3
K5 6 cl 2 on e cl kw on 1 cl e k2 on 0 cl e k1 on 1 e 4 k3 24 IP:Flt3
b
M
L3
cl K on 5 cl e k62 o cl ne w1 on k cl e k 20 on 1 e 14 k3 24 O C cl I - A on M cl e o L o cl ne w1 on o cl e o 35 on 1 e 66 o2 53
0.0
d
ns
K5 62 cl on e kw cl on 1 e k2 0 cl on e k1 14 cl on e k3 24
a
-
+
p-Akt t-Akt p-Erk t-Erk1 p-Stat5 t-Stat5 GAPDH
Figure 2. Single-allelic disruption of FLT3 induces FLT3 haplo-insufficiency in leukemia cells. (a) Quantitative RT-PCR analysis of the FLT3 transcript levels in K562 and OCI-AML3 clones. The relative expression levels normalized to GAPDH are expressed as the means ± SEM. NS, not significant; *P
OPEN
received: 19 March 2015 accepted: 17 November 2015 Published: 16 December 2015
TALENs-mediated gene disruption of FLT3 in leukemia cells: Using genome-editing approach for exploring the molecular basis of gene abnormality Jue Wang1,*, Tongjuan Li1,*, Mi Zhou1, Zheng Hu2, Xiaoxi Zhou1, Shiqiu Zhou1, Na Wang1, Liang Huang1, Lei Zhao1, Yang Cao1, Min Xiao1, Ding Ma2, Pengfei Zhou3, Zhen Shang1 & Jianfeng Zhou1,2 Novel analytic tools are needed to elucidate the molecular basis of leukemia-relevant gene mutations in the post-genome era. We generated isogenic leukemia cell clones in which the FLT3 gene was disrupted in a single allele using TALENs. Isogenic clones with mono-allelic disrupted FLT3 were compared to an isogenic wild-type control clone and parental leukemia cells for transcriptional expression, downstream FLT3 signaling and proliferation capacity. The global gene expression profiles of mutant K562 clones and corresponding wild-type controls were compared using RNA-seq. The transcriptional levels and the ligand-dependent autophosphorylation of FLT3 were decreased in the mutant clones. TALENs-mediated FLT3 haplo-insufficiency impaired cell proliferation and colony formation in vitro. These inhibitory effects were maintained in vivo, improving the survival of NOD/SCID mice transplanted with mutant K562 clones. Cluster analysis revealed that the gene expression pattern of isogenic clones was determined by the FLT3 mutant status rather than the deviation among individual isogenic clones. Differentially expressed genes between the mutant and wild-type clones revealed an activation of nonsense-mediated decay pathway in mutant K562 clones as well as an inhibited FLT3 signaling. Our data support that this genome-editing approach is a robust and generally applicable platform to explore the molecular bases of gene mutations. Acute leukemia (AL) is among the most common forms of hematological malignancies and exhibits a striking heterogeneity in genetic makeup and clinical outcome. Despite striking success in treating some subtypes of AL due to progress in the understanding of cytogenetic/gene mutations and the introduction of corresponding targeted therapies, many genetically high-risk AL subtypes remain refractory to current standard regimens and display poor prognosis, as the current therapeutic strategies are insufficient to cure genetically high-risk AL1,2. The increasing classification/stratification of AL by the World Health Organization (WHO) and National Comprehensive Cancer Network (NCCN) based on genetic orientation highlights the profound impact of cytogenetic/gene mutations on decision-making regarding clinical management. Therefore, there is an urgent need to dissect the molecular basis of AL in these patients and to develop novel treatment strategies. The advances in high-throughput sequencing technologies enabled the characterization of AL from a global genomic perspective in an unbiased and comprehensive manner rather than a candidate gene approach. The identification of highly recurrent gene mutations in AL has provided invaluable preclinical rationales for improved diagnosis, patient stratification and targeted therapy3,4. Nevertheless, although exhaustive landscapes of AL genomes are emerging, elucidating how these recurrent gene mutations are associated with clinical phenotypes and treatment 1
Department of hematology, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Hubei, China. 2Cancer Biology Research Center, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Hubei, China. 3Wuhan YZY Bio-Pharma Co., Ltd., Wuhan, Hubei, China. *These authors contributed equally to this work. Correspondence and requests for materials should be addressed to Z.S. (email: [email protected]) or J.Z. (email: [email protected]) Scientific Reports | 5:18454 | DOI: 10.1038/srep18454
1
www.nature.com/scientificreports/ outcomes has proven difficult. The ability to determine the clinical relevance of these identified individual genetic abnormalities is greatly needed. More importantly, uncovering the molecular events that are central to leukemia biology is a major challenge in the human genome era and is fundamentally important for the future of genetic medicine5. Several approaches have been widely used to explore the molecular mechanism underlying a given gene mutation. Many elegant leukemia animal models have been established using genetically engineered strategies and have provided persuasive in vivo insight into the driving factors of leukemogenesis. In the coming years, genetically engineered animal models will remain among the most reliable tools to study the association between genetic defects and clinical phenotypes, and contribution to the design and development of novel molecular-targeted strategies. Nevertheless, the animal model approach has some intrinsic shortcomings. For example, animal models may not always accurately mimic the leukemia phenotype relevant to the clinical setting6, aside from the time-consuming and expensive process of generating and maintaining a transgenic model. Alternatively, cell lines and primary sample-based models have also been widely used to explore the critical molecular events that lead to tumor cell phenotypes. Several factors have limited the identification of actual molecular mechanisms due to the huge discrepancy in genetic backgrounds among primary samples, the non-physiological levels of transgene overexpression/knockdown and the interference introduced by randomly integrated viral vectors. Therefore, novel analytic tools are urgently needed to address the molecular mechanisms underlying leukemia-relevant gene function in the post-genome era. Transcription activator-like effector (TALE) nucleases (TALENs), an efficient genome editing tool, are artificial fusion proteins containing the catalytic domain of the endonuclease FokI and a designed TALE DNA-binding domain that recognizes a specific DNA sequence7,8. The binding of two separate TALENs to adjacent DNA sequences enables dimerization of FokI and cleavage of the target DNA, introducing site-specific double-strand breaks (DSBs). Subsequently, cellular DNA repair via either homology-directed repair or the non-homologous end joining (NHEJ) pathway is activated9. Thus far, genome-editing technology has been successfully applied to induce myeloid malignancy in normal hematopoietic stem cells in mice10. However, few studies have performed this technique to investigate the molecular events caused by individual gene abnormality in leukemia. In this study, we generated isogenic clones, in two individual leukemia cell lines, by disrupting FMS-like tyrosine kinase 3 (FLT3) gene in a single allele using designed TALENs. The resulting isogenic clones which were only different in the FLT3 mutant status were compared for FLT3 downstream signaling, proliferation capacity and transcriptional expression. Our data strongly support that this genome-editing approach can serve as a robust and generally applicable platform for exploring the molecular basis of a given gene abnormality.
Results
Generation of isogenic leukemia clones carrying disrupted FLT3 in a single allele using designed TALENs. To generate isogenic leukemia clones carrying a disrupted FLT3 juxtamembrane (JM) domain in a
single allele, a pair of TALENs targeting exon14 of FLT3 was designed. The TALEN target site was selected within the sequence of exon14, which encodes the JM domain and is located upstream of the tyrosine kinase domains (TKDs) and the kinase insert (KI) domain, such that insertions or deletions caused by NHEJ could result in disruption of the reading frame or the formation of a premature stop codon. We constructed TALENs composed of 17.5 and 15.5 repeats to cleave a site using a 15 bp spacer according to five computationally derived design guidelines as described previously (Fig. 1a)11. Since the transfection efficiency in K562 cells approached over 90% (Supplementary Fig. S1) under the experimental conditions described in METHODS, the T7E1 mismatch sensitive assay was applied to assess the nuclease activity of the designed TALENs at their intended target in K562 cells. As shown in Fig.1b, T7E1 cleavage products of the expected size were detected on both the electrophoretic trace and the pseudo-gel image, displaying an estimated frequency of NHEJ events of up to 6%. These data demonstrated robust genomic editing capacity of the designed TALENs at the JM domain of endogenous FLT3 in leukemia cells, facilitating subsequent screening of leukemia clones carrying genetically disrupted FLT3. Before establishing isogenic clonal leukemia models, we assessed the basal expression level of FLT3 in a panel of working cell lines representing AML and CML myeloid blast crisis by Western blot (Supplementary Fig. S2). Consistent with previous studies12,13, FLT3 protein levels in cell lines harboring FLT3-ITD mutations such as MV4-11 and MOLM-13 were generally higher than those cell lines with wild-type FLT3, whereas FLT3 wild-type cell lines expressed FLT3 protein at different levels. Based on transfection efficiencies and cell viabilities after nucleofection, we chose three individual cell lines, K562 and OCI-AML3 with the lowest and the highest basal wild-type FLT3 expression respectively, and Jurkat cells with undetectable FLT3 expression as negative control14, to generate isogenic leukemia clones using our designed TALENs. Immunoprecipitation followed by mass spectrometry (Supplementary methods) was performed to address the specificity of FLT3 protein bands in K562 cells (Supplementary Fig. S3 and Supplementary Fig. S4). PCR and Sanger sequencing were conducted to confirm that there was no FLT3-ITD mutation in the three working parental cell lines. The TALEN-nucleofected leukemia cells were seeded at low density in a 96-well plate for recovery and homogeneous expansion. The clonal cell populations were isolated and screened via PCR amplification of genomic DNA and Sanger sequencing to identify indels characteristic of NHEJ (Fig. 1c). In this experiment, among a total of 384 K562 clones screened, 11 clones (~2.86%) contained heterozygous modifications in exon 14, including 5 clones carrying frameshift mutations and 6 clones carrying in-frame deletions in one allele (Fig. 1d). All 5 isogenic K562 clones carrying a frameshift mutation in FLT3 were predicted to contain a premature termination codon, of which three clones (clones k20, k114 and k324) were selected for the subsequent experiments. Similar results were obtained from OCI-AML3 and Jurkat, and 3 clones from OCI-AML3 (clones o35, o166 and o253) and 2 clones from Jurkat (clones j105 and j212) were confirmed with frame shift mutations (Supplementary Table S1)
Scientific Reports | 5:18454 | DOI: 10.1038/srep18454
2
www.nature.com/scientificreports/
a
JM ECD
KI
TM
TKD1
TALEN L exon14
exon13
Protein
TKD2
exon15
Genome
TALEN R primer F
primer R
483 bp
CTTCTACGTTGATTTCAGAGAATATGAATATGATCTCAAATGGGAGTTTCCAAGAGAAA TALEN target GAAGATGCAACTAAAGTCTCTTATACTTATACTAGAGTTTACCCTCAAAGGTTCTCTTT
b
Fraction cleaved:12.5%
bp 500
Estimated %NHEJ:6.46%
Fluorescence Units[x1E3]
20
c 66,455
K562+control 479bp
66,465
Clone k20 wt allele 66,475 66,485
400
10 0 20 10
300
309bp
Clone k20 mut allele
K562+TALENs 479bp 309bp 163bp
0
d
479bp
200 160
163bp
exon 14
100 Clone#
Exon14 sequence
wt k20 k57 k86 k114 k195 k221 k273 k279 k324 k359 k366
CTTCTACGTTGATTTCAGAGAATATGAATATGATCTCAAATGGGAGTTTCCAAGAGAAA CTTCTACGTTGATTTC-------------------TCAAATGGGAGTTTCCAAGAGAAA CTTCTACGTTGATTTCAGA------GAATATGATCTCAAATGGGAGTTTCCAAGAGAAA CTTCTACGTTGATTTCAGA------GAATATGATCTCAAATGGGAGTTTCCAAGAGAAA CTTCTACGTTGATTTCAGAGAATATGA-----ATCTCAAATGGGAGTTTCCAAGAGAAA CTTCTACGTTGATTTCAGA------GAATATGATCTCAAATGGGAGTTTCCAAGAGAAA CTTCTACGTTGATTTCAGAGAATATGAA---------AAATGGGAGTTTCCAAGAGAAA CTTCTACGTTGATTTCAGAGAATATG------------AATGGGAGTTTCCAAGAGAAA CTTCTACGTTGATTTCAGA------GAATATGATCTCAAATGGGAGTTTCCAAGAGAAA CTTCTACGTTGATTTCAGAGAATATG-----------AAATGGGAGTTTCCAAGAGAAA CTTCTACGTTGATTTCAGAGAATATCAATTATGATCTCAAATGGGAGTTTCCAAGAGAA CTTCTACGTTGATTTCAGAGAAT----------------ATGGGAGTTTCCAAGAGAAA
Frame shift 19 6 6 5 6 9 12 6 11 1 16
NHEJ events ( 11/384=~2.86% )
Figure 1. TALEN-mediated gene disruption of FLT3 in the JM domain in leukemia cells. (a) Schematic overview of the strategy to disrupt the JM domain of FLT3. The three exons that encode the JM domain, the TALEN pair used (TALEN L and TALEN R) and the primer pair (Primer F and Primer R) for PCR and DNA sequencing to screen the mutant clones were shown. The boxes indicate the TALEN binding sequences in the lower diagram in grey. (b) A T7 Endonuclease I assay presented as an electrophoretic trace (the left upper and lower panels) and a pseudo-gel image (the right panel). The blue-colored trace and the lane labeled with a blue bar at the top correspond to K562 cells transfected with the empty TALEN expression vector as a negative control. The red-colored trace and the lane labeled with a red bar at the top correspond to K562 cells transfected with the TALEN pair. The red arrows denote T7EI cleavage products of the appropriate size, and the single black arrow denotes the original uncleaved PCR product in both the electrophoretic trace and the pseudo-gel image. The internal lane standards were labeled with a black bar. (c) Representative sequencing results of targeted K562 clones displaying either the wild-type or targeted mutant allele. The original base numbers in the genomic region of the wild-type FLT3 gene were shown. The dotted lines denote the 19bp deletion in the mutant allele. The colored bars below each chromatogram indicate the predicted proteins. The black bar with an asterisk represents the premature termination codon. (d) Summary of Sanger sequencing results of positive isogenic mutant clones derived from the K562 cell line transfected with TALENs targeting FLT3 exon14. The TALEN binding sequences are shown in bold letters in the wild-type clone. The deletions are indicated by dashes, and insertion or base substitution is denoted by a black triangle or an asterisk, respectively. ECD, extracellular domain; TM, transmembrane domain; JM, juxtamembrane domain; TKD, tyrosine kinase domain; KI, kinase insert; TALEN, transcription activator-like effector nuclease; WT, wild-type; Mut, mutant; Bp, base pair; NHEJ, non-homologous end joining. Scientific Reports | 5:18454 | DOI: 10.1038/srep18454
3
www.nature.com/scientificreports/
Clone analyzed
Putative off-target sites examined
Mutated sites detected (reference sequence)
Comparison to parental K562 sequence
Clone kw1
20
2
Identical (20/20)
Clone k20
20
2
Identical (20/20)
Clone k114
20
2
Identical (20/20)
Clone k324
20
2
Identical (20/20)
Table 1. Summary of the off-target analysis to isogenic clones. The 20 most possible off-target sites identified using the TALE-NT 2.0 Paired Target Finder were sequenced and aligned with either reference sequence (human reference genome, NCBI Build 36.1) or parental K562 sequence.
after screening 480 OCI-AML3 clones and 300 Jurkat clones, respectively. Three wild-type clones derived from the same screening procedure for each cell line were randomly selected as negative controls. To assess the off-target gene effects of the designed TALENs, Paired Target Finder was employed to search for potential off-target sites. The genomic region containing the 20 most likely off-target candidate sites was PCR-amplified and subjected to sequencing analysis. Mutations within two candidate sites were identified in all isogenic K562 subclones examined; however, the parental K562 cells were confirmed to carry the identical mutations (Table 1). Therefore, no off-target events caused by the designed TALENs were detected.
Single-allelic disruption of FLT3 induces FLT3 haplo-insufficiency in Leukemia cells. To deter-
mine whether the premature termination codon caused by NHEJ mutation in the mutant clones of each cell lines induces FLT3 haplo-insufficiency, we first examined the FLT3 mRNA transcript levels in the parental cell line and its isogenic clones via real-time RT-PCR. As expected, significantly lower transcript levels of FLT3 were detected in the mutant clones compared to those of wild-type clones or the parental cell population for both K562 and OCI-AML3 (Fig.2a), implying that premature stop codon-mediated mRNA decay might occur in the mutant FLT3 clones. Next, we sought to determine whether the NHEJ mutation induced a reduced level of FLT3 autophosphorylation upon exposure to treatment of the isogenic mutant clones with FLT3 ligand (FL). Since the K562 cells express the FLT3 protein at a very low level, immunoblotting was performed following immunoprecipitation with FLT3 antibody to examine the potential FL-dependent activity of FLT3 in the K562 clones. As shown in Fig.2b, western blot analyses revealed substantially attenuated FL-dependent autophosphorylation of FLT3 in the mutant clones compared to the wild-type K562 clones, although the reduction of total FLT3 protein level was not significant after immunoprecipitation. The target genes downstream of the FLT3 signaling pathway were further examined via Western blot. Although consistent activation of FLT3 target genes was detected in a FL-dependent manner in all K562 clones, the FL-induced phosphorylation level was substantially reduced in the mutant K562 clones compared to their wild-type counterparts (Fig. 2c). Because K562 cells carry the oncogenic bcr/abl fusion gene, which is also a potent activator of the downstream pathways described above15, we sought to determine whether the TALENs exerted off-target effects on the bcr/abl protein. As shown in Fig. 2d, neither the basal nor phosphorylated level of bcr/abl was affected by the TALENs procedure, indicating that FLT3 haplo-insufficiency was specifically caused by the introduced mutations. A set of similar immunoblotting assays was performed to validate the TALENs-mediated FLT3 haplo-insufficiency for OCI-AML3 clones (Fig. 2e). As results, a significant down-regulation of FLT3 protein expression and ligand-dependent FLT3 autophosphorylation in the mutant OCI-AML3 clones was easily detected using standard immunoblotting, whereas similar trends of weakened Akt, Erk and Stat5 signaling were observed. However, no such alterations regarding the FLT3 protein expression or downstream signaling could be detected in the negative control Jurkat cells and its isogenic clones (Supplementary Fig. S5). To confirm that the changes in cellular signaling were caused by FLT3 gene disruption, siRNA knockdown was performed in K562, OCI-AML3 and THP-1 (Supplementary methods). As shown in Supplementary Fig. S6, similar alterations in FLT3 signaling occurred 48 hours after siRNA transfection. Collectively, these results demonstrate that TALEN-mediated disruption of exon14 of FLT3 induces FLT3 haplo-insufficiency as indicated by reduced FLT3 autophosphorylation and substantially attenuated FLT3 signaling.
TALEN-mediated FLT3 haplo-insufficiency impairs cell proliferation and colony forming capacity in vitro. It is well established that one of the most prominent characteristics of activated FLT3 signaling is the
promotion of cell proliferation16. Therefore, we sought to investigate the biologic effect of TALEN-mediated FLT3 haplo-insufficiency on leukemia cell proliferation in vitro. Based on the CCK8 (Fig. 3a) and CFSE (Fig. 3b) assays in both K562 and OCI-AML3 cells, the proliferation capacity in the mutant FLT3 clones was consistently reduced to nearly 50% or less than those of the corresponding wild-type FLT3 clones, while no inhibition on the proliferation capacity could be observed in the mutant Jurkat clones compared to their wild-type counterparts (Supplementary Fig. S7). In the same experiments, no significant difference in cell proliferation capacity was detected between the isogenic mutant clones (P > 0.05). It was also to our surprise that even in K562 cells that express FLT3 at a very low level, the TALEN-mediated FLT3 haplo-insufficiency resulted in significantly impaired proliferation capacity. To rule out the possibility that TALEN-mediated growth inhibition might be an artefact caused by multiple experimental procedures, we generated and sequence-verified an artificial heterozygous FLT3-ITD model (clone KI-H6) in K562 using the same pair of TALENs by homologous directed repair adopting similar protocols as previously reported (Supplementary Figure S8)17. By comparing growth capacities of three clones with different FLT3 mutant status (FLT3-wt/wt, wt/-, wt/ITD), we found that the FLT3-ITD knock-in clone (KI-H6) display similar proliferative patterns to that of FLT3 wt/wt clones in the presence of serum (Supplementary Figure S8). The findings support that the growth inhibition on FLT3 wt/- clones was specifically due to FLT3 gene disruption Scientific Reports | 5:18454 | DOI: 10.1038/srep18454
4
www.nature.com/scientificreports/ RT-PCR FLT3 1.5
1.5
ns
1.0
1.0
0.5
0.5
p-abl
Bcr-abl β-actin
0.0
Flt3 ligand -
+
-
p-Akt t-Akt p-Erk t-Erk p-Stat5 t-Stat5 GAPDH
+
-
+
-
+
-
-
+
o2 53
o1 66
o3 5
1
cl on e
ow +
cl on e
+
cl on e
Flt3 ligand -
cl on e
O C I-A
e
-
-
+
+
p-Flt3
k3 24
Flt3
cl
on e
k1 14
on e cl
on e cl
on e cl
K5 62
kw
c
k2 0
1
p-Flt3 Flt3
K5 6 cl 2 on e cl kw on 1 cl e k2 on 0 cl e k1 on 1 e 4 k3 24 IP:Flt3
b
M
L3
cl K on 5 cl e k62 o cl ne w1 on k cl e k 20 on 1 e 14 k3 24 O C cl I - A on M cl e o L o cl ne w1 on o cl e o 35 on 1 e 66 o2 53
0.0
d
ns
K5 62 cl on e kw cl on 1 e k2 0 cl on e k1 14 cl on e k3 24
a
-
+
p-Akt t-Akt p-Erk t-Erk1 p-Stat5 t-Stat5 GAPDH
Figure 2. Single-allelic disruption of FLT3 induces FLT3 haplo-insufficiency in leukemia cells. (a) Quantitative RT-PCR analysis of the FLT3 transcript levels in K562 and OCI-AML3 clones. The relative expression levels normalized to GAPDH are expressed as the means ± SEM. NS, not significant; *P
